Compressibility correction to the k-$\omega$ turbulence model that considers the wall-cooling effect

TL;DR

This paper introduces a modified k-omega turbulence model that accounts for wall cooling effects in supersonic and hypersonic flows, improving near-wall velocity and temperature predictions by aligning with the universal wall law.

Contribution

A simple modification to the k-omega model based on dimensional analysis that enhances accuracy in compressible turbulent boundary layers with wall cooling effects.

Findings

Improves velocity profile predictions in compressible turbulent flows.

Enhances temperature predictions in hypersonic boundary layers.

Maintains the original model's robustness in pressure gradient conditions.

Abstract

In supersonic and hypersonic flows, the near-wall density variation due to wall cooling poses a challenge for accurately predicting the near-wall velocity and temperature profiles using classical eddy viscosity turbulence models. Compressible turbulent boundary layers are known to follow the universal wall law via semi-local transformation. However, developing a turbulence model that predicts a velocity profile, which, via semi-local transformation, follows the universal wall law, remains challenging. The current paper builds upon Danis-Durbin's practice of modifying the $\omega$ equation and proposes a simple modification to the $k-\omega$ two-equation model. The formulation of the proposed modification involves dimensional analysis and the proper selection of the local length scale. The newly introduced modification is used to modify the slope of the velocity profile starting from the viscous layer to above. It recovers a semi-local scaling of turbulent kinetic energy, viscosity, and eddy frequency, then achieves a very decent correction of the velocity profile in compressible turbulent channel flows, satisfying the universal wall law after applying Trettel \& Larsson's transformation. The proposed new $k-\omega$ model can also improve the velocity and temperature predictions in strongly wall-cooled zero-pressure-gradient hypersonic turbulent boundary layers, compared to the original $k-\omega$ model. Validation using the favorable and adverse pressure gradient boundary layers suggests that the model does not impose a negative effect on the original $k-\omega$ model.